]>
https://gitweb.dealii.org/ - dealii.git/log
Martin Kronbichler [Fri, 8 Feb 2019 15:28:21 +0000 (16:28 +0100)]
Adjust Chebyshev degree in loop to convention in literature.
Jean-Paul Pelteret [Fri, 8 Feb 2019 07:26:19 +0000 (08:26 +0100)]
Merge pull request #7704 from masterleinad/unrestrict_scalapack_copy_to_from
Unrestrict ScaLAPACKMatrix::copy_to/from
Wolfgang Bangerth [Fri, 8 Feb 2019 04:58:50 +0000 (21:58 -0700)]
Merge pull request #7645 from marcfehling/hp-dofhandler-transfer
Automatic transfer of active fe indices.
Marc Fehling [Tue, 18 Dec 2018 14:23:15 +0000 (15:23 +0100)]
Automatic transfer of active fe indices during refinement and serialization.
Daniel Arndt [Thu, 7 Feb 2019 17:04:32 +0000 (18:04 +0100)]
Unrestrict ScaLAPACKMatrix<NumberType>::copy_to/from
Daniel Arndt [Thu, 7 Feb 2019 16:48:40 +0000 (17:48 +0100)]
Merge pull request #7697 from tjhei/ci_jenkins_marker
[CI]: add Jenkins marking
Martin Kronbichler [Thu, 7 Feb 2019 09:25:38 +0000 (10:25 +0100)]
Merge pull request #7700 from Rombur/complex
Do not use I in header files due to conflict with <complex.h>
Martin Kronbichler [Thu, 7 Feb 2019 09:24:50 +0000 (10:24 +0100)]
Merge pull request #7678 from masterleinad/fix_tests_p4est
Fix p4est test
Martin Kronbichler [Thu, 7 Feb 2019 09:23:09 +0000 (10:23 +0100)]
Merge pull request #7679 from masterleinad/fix_tbb_nvcc_wrapper
Fix building TBB with nvcc_wrapper
Bruno Turcksin [Wed, 6 Feb 2019 22:00:24 +0000 (17:00 -0500)]
Do not use I in header files due to conflict with <complex.h>
The C version of <complex.h> defines the macro I. So when a third-party
library such as lapacke include <complex.h> is included before the
header the variable I is expanded which trips the compiler.
Wolfgang Bangerth [Wed, 6 Feb 2019 16:31:47 +0000 (09:31 -0700)]
Merge pull request #7699 from davydden/feature/block_indices_global_to_local
use binary search in BlockIndices::global_to_local()
Timo Heister [Wed, 6 Feb 2019 16:23:13 +0000 (09:23 -0700)]
Merge pull request #7680 from masterleinad/document_MPI_minimal
Document the minimal supported MPI version
Denis Davydov [Wed, 6 Feb 2019 13:18:56 +0000 (14:18 +0100)]
use binary search in BlockIndices::global_to_local()
Martin Kronbichler [Wed, 6 Feb 2019 09:47:27 +0000 (10:47 +0100)]
Merge pull request #7687 from stefanmeggendorfer/mappingqgenericfix
MappingQGeneric: use of MatrixFree only for higher order elements
Martin Kronbichler [Wed, 6 Feb 2019 09:46:58 +0000 (10:46 +0100)]
Merge pull request #7688 from masterleinad/more_output_matrix_free_dof_info
More output files for matrix_free/dof_info_01/2
Timo Heister [Tue, 5 Feb 2019 23:08:43 +0000 (16:08 -0700)]
set agent
Timo Heister [Tue, 5 Feb 2019 23:02:13 +0000 (16:02 -0700)]
[CI]: add Jenkins.mark
Add a separate Jenkins pipeline that runs in a few seconds to set the
commit status to pending. This is needed as Jenkins only starts setting
a commit status when the normal jobs start running and not when they are
in the queue. This can lead to a) accidental merges before we test
things and b) repeated /rebuild triggers.
Daniel Arndt [Tue, 5 Feb 2019 09:36:40 +0000 (10:36 +0100)]
Merge pull request #7683 from drwells/cleanup-vector-thread-partitioner
Cleanup the Vector::thread_loop_partitioner logic.
Daniel Arndt [Tue, 5 Feb 2019 09:05:38 +0000 (10:05 +0100)]
Merge pull request #7691 from gassmoeller/fix_xdmf_entry_serialization
Fix serialization of XDMFEntry class
Daniel Arndt [Tue, 5 Feb 2019 08:47:55 +0000 (09:47 +0100)]
Merge pull request #7686 from bangerth/doc-update-99
Update documentation of some VectorTools functions.
Wolfgang Bangerth [Mon, 4 Feb 2019 14:51:52 +0000 (07:51 -0700)]
Update documentation of some VectorTools functions.
David Wells [Mon, 4 Feb 2019 03:35:36 +0000 (22:35 -0500)]
Cleanup the Vector::thread_loop_partitioner logic.
It was previously possible to, in one function call, either set up the
partitioner multiple times or for inappropriately small vectors. This
commit cleans up the way we handle the partitioner in multiple places so
that, when possible, vectors share partitioners and do not set up
partitioners if they are too small.
Rene Gassmoeller [Mon, 4 Feb 2019 21:12:12 +0000 (13:12 -0800)]
Fix serialization of XDMFEntry class
Wolfgang Bangerth [Mon, 4 Feb 2019 20:47:50 +0000 (13:47 -0700)]
Merge pull request #7689 from tamiko/fix_example_step_4
example/step-4: Return the correct return value
Matthias Maier [Mon, 4 Feb 2019 16:55:39 +0000 (10:55 -0600)]
example/step-4: Return the correct return value
This commit reverts a stray change that should have probably never made
it into
0e12f6f87a0b84fadd8df629f55da42a8494ce8f
Daniel Arndt [Mon, 4 Feb 2019 16:12:06 +0000 (17:12 +0100)]
More output files for matrix_free/dof_info_01/2
Denis Davydov [Mon, 4 Feb 2019 14:53:30 +0000 (15:53 +0100)]
Merge pull request #7681 from tamiko/fix_openmp_simd_detection
CMake: Bugfix: Guard DEAL_II_HAVE_OPENMP_SIMD test
Stefan Meggendorfer [Mon, 4 Feb 2019 13:56:57 +0000 (14:56 +0100)]
MappingQGeneric: use of MatrixFree only for higher order elements
Daniel Arndt [Mon, 4 Feb 2019 09:13:14 +0000 (10:13 +0100)]
Merge pull request #7677 from drwells/avoid-deprecated-mpi
Avoid calling a deprecated MPI function.
Daniel Arndt [Mon, 4 Feb 2019 01:08:20 +0000 (02:08 +0100)]
Fix MPI support for older versions
Matthias Maier [Sun, 3 Feb 2019 18:06:59 +0000 (12:06 -0600)]
CMake: Bugfix: Guard DEAL_II_HAVE_OPENMP_SIMD test
We also have to guard the DEAL_II_HAVE_OPENMP_SIMD test with
DEAL_II_ALLOW_PLATFORM_INTROSPECTION. If the latter is set to false our
cmake configuration does not attempt to determine any specifics of the
unerlying CPU or architecture (and instead relies on the user specifying
everything by hand). This is necessary for scenarios where deal.II gets
compiled on a different machine than the one it will be run on
eventually.
In reference to #7663.
Martin Kronbichler [Sun, 3 Feb 2019 16:27:22 +0000 (17:27 +0100)]
Merge pull request #7641 from davydden/feature/fe_eval_dofs
add internal::MatrixFreeFunctions::DoFInfo::get_dof_indices()
Daniel Arndt [Sun, 3 Feb 2019 13:04:33 +0000 (14:04 +0100)]
Remove unused variables
Daniel Arndt [Sun, 3 Feb 2019 11:06:36 +0000 (12:06 +0100)]
Fix building TBB with nvcc_wrapper
Daniel Arndt [Sat, 2 Feb 2019 23:44:45 +0000 (00:44 +0100)]
Merge pull request #7674 from bangerth/simplify
Simplify some code.
Daniel Arndt [Sat, 2 Feb 2019 17:45:12 +0000 (18:45 +0100)]
Document the minimal supported MPI version
Daniel Arndt [Sat, 2 Feb 2019 16:11:30 +0000 (17:11 +0100)]
Merge pull request #7675 from dangars/unused_variable_warning
Remove unused variable warning
Daniel Arndt [Sat, 2 Feb 2019 12:05:31 +0000 (13:05 +0100)]
Fix p4est test
Wolfgang Bangerth [Fri, 1 Feb 2019 15:04:58 +0000 (08:04 -0700)]
Simplify some code.
David Wells [Sat, 2 Feb 2019 15:00:38 +0000 (10:00 -0500)]
Avoid calling a deprecated MPI function.
This was deprecated a long time ago (1996) and is not present in the
latest version of openMPI (4.0): see
https://www.open-mpi.org/faq/?category=mpi-removed
Credit goes to Pratik Nayak for finding this issue.
Daniel Garcia-Sanchez [Sat, 2 Feb 2019 13:26:44 +0000 (14:26 +0100)]
Remove unused variable warning
Daniel Arndt [Sat, 2 Feb 2019 11:02:07 +0000 (12:02 +0100)]
Merge pull request #7666 from gassmoeller/ensure_particle_property_initialization
Ensure particle property initialization
Rene Gassmoeller [Thu, 31 Jan 2019 18:24:06 +0000 (10:24 -0800)]
Ensure particle property initialization
Daniel Arndt [Fri, 1 Feb 2019 17:56:33 +0000 (18:56 +0100)]
Merge pull request #7672 from masterleinad/fix_msvc
Fix compiling with MSVC
Wolfgang Bangerth [Fri, 1 Feb 2019 16:30:25 +0000 (09:30 -0700)]
Merge pull request #7667 from gassmoeller/fix_serialization_of_variable_data
Fix serialization of variable data attachements
Daniel Arndt [Fri, 1 Feb 2019 13:12:29 +0000 (14:12 +0100)]
Fix compiling with MSVC
Jean-Paul Pelteret [Fri, 1 Feb 2019 11:42:13 +0000 (12:42 +0100)]
Merge pull request #7669 from masterleinad/fix_old_clang
Be explicit about initialization of static const Tensor variables
Daniel Arndt [Thu, 31 Jan 2019 22:48:28 +0000 (23:48 +0100)]
Merge pull request #7651 from tjhei/clang_tidy_rework
[CI] introduce clang-tidy check
Daniel Arndt [Thu, 31 Jan 2019 22:47:38 +0000 (23:47 +0100)]
Be explicit about initialization of static const Tensor variables
Rene Gassmoeller [Thu, 31 Jan 2019 19:45:03 +0000 (11:45 -0800)]
Fix serialization of variable data attachements
Timo Heister [Tue, 29 Jan 2019 19:44:47 +0000 (12:44 -0700)]
convert for loops
Timo Heister [Mon, 28 Jan 2019 21:17:53 +0000 (14:17 -0700)]
[CI] introduce clang-tidy check
- remove -fix
- allow bundled/ by including two .clang-tidy files
- add jenkins script
Wolfgang Bangerth [Thu, 31 Jan 2019 14:19:35 +0000 (07:19 -0700)]
Merge pull request #7637 from masterleinad/fix_gmsh_bid
Parse boundary_ids in GMSH-4 format correctly
Daniel Arndt [Thu, 31 Jan 2019 12:17:56 +0000 (13:17 +0100)]
Merge pull request #7661 from gassmoeller/fix_particle_property_serialization_bug
Fix invalid memory access in particle properties after transfer
Jean-Paul Pelteret [Thu, 31 Jan 2019 08:18:54 +0000 (09:18 +0100)]
Merge pull request #7368 from ginkgo-project/ginkgo-wrapper
Ginkgo wrapper for deal.ii
Daniel Arndt [Thu, 31 Jan 2019 07:47:07 +0000 (08:47 +0100)]
Merge pull request #7654 from GivAlz/GeneralCell
Generalizing GridGenerator::general_cell to tria<dim,spacedim>
Rene Gassmoeller [Thu, 31 Jan 2019 00:42:20 +0000 (16:42 -0800)]
Fix invalid memory access in particle properties after transfer
Jean-Paul Pelteret [Wed, 30 Jan 2019 21:48:35 +0000 (22:48 +0100)]
Merge pull request #7509 from jppelteret/ad-drivers_03
Refactor the AD drivers classes
Daniel Arndt [Wed, 30 Jan 2019 21:44:08 +0000 (22:44 +0100)]
Merge pull request #7655 from masterleinad/fix_cuda_vectorization
Disable SIMD vectorization in CUDA device code
Denis Davydov [Fri, 25 Jan 2019 21:18:17 +0000 (22:18 +0100)]
Add internal::MatrixFreeFunctions::DoFInfo::get_dof_indices_on_cell_batch()
Pratik Nayak [Wed, 30 Jan 2019 15:20:39 +0000 (16:20 +0100)]
Add changelog entry and fix a typo.
Jean-Paul Pelteret [Wed, 30 Jan 2019 13:20:29 +0000 (14:20 +0100)]
Improve documentation of some functions
Bruno Turcksin [Wed, 30 Jan 2019 13:23:28 +0000 (08:23 -0500)]
Merge pull request #7656 from masterleinad/add_cuda_header
Add missing headers in CUDA header file
Pratik Nayak [Wed, 30 Jan 2019 11:50:06 +0000 (12:50 +0100)]
Add the ginkgo source and include files.
Pratik Nayak [Wed, 30 Jan 2019 11:49:36 +0000 (12:49 +0100)]
Add initial tests for ginkgo
Pratik Nayak [Wed, 30 Jan 2019 11:49:04 +0000 (12:49 +0100)]
Add documentation to readme.html and ginkgo.html files
Pratik Nayak [Wed, 30 Jan 2019 11:47:50 +0000 (12:47 +0100)]
Detect Ginkgo and compile the ginkgo files: CMake.
Giovanni Alzetta [Tue, 29 Jan 2019 15:29:15 +0000 (16:29 +0100)]
Generalizing GridGenerator::general_cell to tria<dim,spacedim>
Wolfgang Bangerth [Wed, 30 Jan 2019 03:10:43 +0000 (20:10 -0700)]
Merge pull request #7633 from drwells/vector-use-aligned-vector
Use AlignedVector in Vector.
Daniel Arndt [Tue, 29 Jan 2019 22:48:08 +0000 (23:48 +0100)]
Add missing headers in CUDA header file
Daniel Arndt [Tue, 29 Jan 2019 22:23:50 +0000 (23:23 +0100)]
Disable SIMD vectorization in CUDA device code
Wolfgang Bangerth [Tue, 29 Jan 2019 16:12:18 +0000 (09:12 -0700)]
Merge pull request #7652 from drwells/inline-subscriptor-ctors
Mark a few Subscriptor functions as inline.
David Wells [Sun, 27 Jan 2019 00:23:24 +0000 (19:23 -0500)]
Update the .gdbinit script.
David Wells [Mon, 21 Jan 2019 03:32:32 +0000 (22:32 -0500)]
Use AlignedVector to manage Vector's memory.
Vector already requires that the underlying buffer be aligned and uses
exactly the same function (posix_memalign) as AlignedVector to do it: we
can simplify the implementation of Vector a lot by just using an
AlignedVector to manage the data instead.
Wolfgang Bangerth [Tue, 29 Jan 2019 13:13:11 +0000 (06:13 -0700)]
Merge pull request #7583 from GivAlz/TriangulationCovering
Add "covering rtree" to GridTools::Cache
Wolfgang Bangerth [Tue, 29 Jan 2019 13:12:48 +0000 (06:12 -0700)]
Merge pull request #7653 from masterleinad/more_clang_tidy_fixes_2
Fix more clang-tidy findings
Giovanni Alzetta [Wed, 9 Jan 2019 21:13:56 +0000 (22:13 +0100)]
Added covering rtree and flag to GridTools::Cache
Daniel Arndt [Tue, 29 Jan 2019 09:59:11 +0000 (10:59 +0100)]
performance-unnecessary-value-param
Daniel Arndt [Tue, 29 Jan 2019 09:58:02 +0000 (10:58 +0100)]
modernize-use-nullptr
Daniel Arndt [Tue, 29 Jan 2019 09:57:27 +0000 (10:57 +0100)]
modernize-use-equals-default
Daniel Arndt [Tue, 29 Jan 2019 09:52:18 +0000 (10:52 +0100)]
modernize-return-braced-init-list
Daniel Arndt [Tue, 29 Jan 2019 08:53:54 +0000 (09:53 +0100)]
Merge pull request #7648 from tjhei/jenkins_osx
[CI] test on OSX
Daniel Arndt [Tue, 29 Jan 2019 01:01:52 +0000 (02:01 +0100)]
Move some declarations around and document tag_maps
David Wells [Mon, 28 Jan 2019 21:51:31 +0000 (16:51 -0500)]
Mark a few Subscriptor functions as inline.
These are simple enough that they should live in the header. This will
allow classes inheriting from Subscriptor to potentially inline some
more constructors too.
Timo Heister [Thu, 17 Jan 2019 00:43:46 +0000 (17:43 -0700)]
[CI] test on OSX
- add a separate Jenkinsfile for OSX
- compile deal.II using clang on OSX
Jean-Paul Pelteret [Mon, 28 Jan 2019 20:07:00 +0000 (21:07 +0100)]
Merge pull request #7650 from masterleinad/n_array_elemnts_outside
Declare VectorizedArray::n_array_elements outside the class in the general case
Wolfgang Bangerth [Mon, 28 Jan 2019 15:11:09 +0000 (08:11 -0700)]
Merge pull request #7647 from drwells/test-use-zoltan
Explicitly use Zoltan in two tests.
Daniel Arndt [Mon, 28 Jan 2019 13:15:57 +0000 (14:15 +0100)]
Declare VectorizedArray::n_array_elemnts outside the class in the general case
Bruno Turcksin [Sun, 27 Jan 2019 22:46:57 +0000 (17:46 -0500)]
Merge pull request #7643 from masterleinad/fix_trilinoswrappers_precondition_muelu
Fix TrilinosWrappers::PreconditionMueLu
David Wells [Sun, 27 Jan 2019 21:48:17 +0000 (16:48 -0500)]
Explicitly use Zoltan in two tests.
This doesn't matter in 05 (since that test only uses one processor) but
the output file states that we use Zoltan, so we should be
consistent.
This is needed in 07 since the default third argument to
partition_triangulation is metis, but the output file states that we use
Zoltan.
David Wells [Sun, 27 Jan 2019 20:25:37 +0000 (15:25 -0500)]
Merge pull request #7644 from masterleinad/avoid_builtin_expect
Don't use __builtin_expect in Utilities::pow
Daniel Arndt [Sun, 27 Jan 2019 10:15:31 +0000 (11:15 +0100)]
Add more comments
Daniel Arndt [Sat, 26 Jan 2019 16:35:22 +0000 (17:35 +0100)]
Use default destructor and avoid constructor to be defaulted
Daniel Arndt [Sat, 26 Jan 2019 15:54:16 +0000 (16:54 +0100)]
Rename variables using snake_case style
Daniel Arndt [Sat, 26 Jan 2019 15:13:16 +0000 (16:13 +0100)]
Don't use __builtin_expect in Utilities::pow
Daniel Arndt [Sat, 26 Jan 2019 14:34:33 +0000 (15:34 +0100)]
Fix TrilinosWrappers::PreconditionMueLu
Wolfgang Bangerth [Fri, 25 Jan 2019 21:40:56 +0000 (14:40 -0700)]
Merge pull request #7639 from tamiko/fix_step_1
examples/step-1: Remove explicit reset_manifold
Wolfgang Bangerth [Fri, 25 Jan 2019 21:40:26 +0000 (14:40 -0700)]
Merge pull request #7222 from davydden/feature/qr
QR classes for solvers
Matthias Maier [Fri, 25 Jan 2019 19:26:43 +0000 (13:26 -0600)]
examples/step-1: Remove explicit reset_manifold
We have restructured our internal manifold handling and the explicit
call to reset_manifold is not necessary any more. (Indeed the
description why this reset is necessary is now factually wrong.) Let's
not confuse the unsuspecting user and remove this part.
Thanks to Mateusz Marciniak for pointing this out.
Daniel Arndt [Fri, 25 Jan 2019 14:20:49 +0000 (15:20 +0100)]
Add GMSH-4 tests
In the beginning the Universe was created. This has made a lot of
people very angry and has been widely regarded as a bad move.
Douglas Adams
Typeset in
Trocchi and Trocchi Bold Sans Serif.